Examples

Complete, runnable walkthroughs — from your first push to autopilot goals, scenarios, and alerts.

Every example on these pages is a complete story built with the official SDKs — run the blocks top to bottom in a notebook or a script and they work. They start trivial and build up to the platform's most powerful surfaces.

All you need is an API token:

pip install predictai        # Python
npm install @predictai/sdk   # Node

Start here

Hello, predictAI

The smallest possible program: push values, read them back, and see your data start to connect. Five minutes, five calls.

A forecast with zero training

Deploy a pretrained foundation model zero-shot and get a probabilistic forecast — no training run, no waiting.

CSV to live forecast

The full production loop: load history from a CSV, define a segment, train a model, deploy the winner, and serve forecasts over HTTP.

Go deeper

Race a model pool

Put a custom LSTM and a foundation model in the same pipeline, let every run train both, and deploy whichever wins.

A goal on autopilot

Name an outcome and let the platform find its drivers, train champions, and keep a forecast live — then ask it why.

What-if scenarios

Move a driver, replay an event, and watch the forecast respond — counterfactuals served by the live champion.

Alerts that watch the forecast

Put a condition on the future — get told the moment the forecast crosses it, on the channels you choose.

An anomaly goal, end to end

The capstone: one CSV upload, an anomaly goal, alerts — and every step watched live over WebSockets instead of polling.

The samples use Python and Node tabs — pick your language once and every page on the site follows. IDs (ws_…, segment and pipeline UIDs) are placeholders: each example tells you where its inputs come from.